_Gitana_. Such results explain themselves when we reflect that at these
great speeds the _Gitana_ sinks to such a degree that the afterside
planks are at the level of the water, while the Pictet model rises
simultaneously fore and aft, thus considerably diminishing the submerged
section.

With low or moderate speeds there is a perceptible equality between the
theoretic curve and the curve of the fast boat; but, starting from 16
kilometers, the stress diminishes. The greater does the speed become,
the more considerable is the diminution in stress; and, starting from a
certain speed, the rise of the boat is such as to diminish its absolute
tractive stress--a fact of prime importance established by theory and
confirmed by experiment.

The curves in Fig. 4 show the power in horses necessary to effect
progression at different speeds. The curve, A, has reference to an
ordinary boat that preserves its water lines constant, and the curve,
B, to a swift boat of the same tonnage. Up to 16 kilometers, the swift
vessel presents no advantage; but beyond that speed, the advantage
becomes marked, and, at a speed of 27 kilometers, the power to be
expended is no more than half that which corresponds to the same speed
for an ordinary boat.

The water escapes in a thin and even sheet as soon as the tractive
stress exceeds 2,000 kilogrammes; and the intensity and size of
the eddies from the boat sensibly diminish in measure as the speed
increases.

The interesting experiments made by Mr. Pictet seem, then to clearly
establish the fact that the forms deduced by calculation are favorable